It doesn’t have to be ‘immediately’ useful, it doesn’t have to be useful at all. But, if we are to care about truth, we care about truth FOR some reason. The question “why do I care about Truth” presupposes that it might be possible to answer “I don’t.” So it is not intrinsically necessary that we care about Truth, so it is up to us to justify our desire for it.

I think truth is useful to many people. You want your romantic partner to truly love you, you care about the truth of their feelings. Many examples where we care about truth. So then how do we get truth? You’re right, an inexact but working model of it is the best we can do. Thats what I was also getting at with my idea that we don’t care so much about absolute Truth, but only subjective truth.

I’m not proposing an idealist theory of Truth. Your case of the red ball has an important nuance. It is not about the present indeterminacy of the red ball being there or not. This is something that CAN be checked, it is a falsifiable theory. If you were to add as a premise that the box is indestructible and that it is impossible to verify whether there is a red ball or not, then we have an issue where we have to banish the possibility of objective Truth.

There might in fact be an objective Truth, the ball is there or it isn’t, but what good is it to us if its impossible to have the truth one way or another? If I correctly guess that the red ball is in there (assuming the objective Truth is the red ball is there), I will never be able to benefit from the objectivity of this truth, because for me it will always be doubtful, so it is inevitably reduced to the level of belief. So what good does it do me what the objective truth is? Even if I am holding the truth in my hand, I don’t benefit from it.

This indestructible box example is a better metaphor for your question of the consciousness of another being (AI, or even another human), because as far as an outsider is concerned, we can only make educated guesses based on intuition about the existence of a consciousness in another mind. Strictly speaking, you have never directly seen concrete evidence of another person’s mind, and so based on a theory of 100% certainty, you might as well be solipsistic. But we operate without having to know absolute truth. I operate on the basis of a very strong intuition that other minds are conscious, without a certainty about it. I have banished the need to know the objective truth here and allowed myself to be satisfied with an uncertain, but accessible and functional, subjective truth.

There are some truths that are inaccessible. I referenced Kant and Gödel’s incompleteness theorem for that reason. How did the universe start? Do I have free will? Is there a God? What is Beauty? What is my purpose? These are not Truths that some other human (nor any other creature, unless they are God) will be able to stumble across, they are fundamentally unknowable.

I mentioned Ivan Karamazov’s “Even If” because it is a good reason to assimilate truth with what can be known of it.

What value is there in a truth that can never be known? Is there a good reason to maintain a notion of truth that is inaccessible? It becomes totally useless to us at that point

I think there might be some nuance that can be added in your second paragraph.

What we know to be true is not the same as what is true, I agree. However, in answering the question of what is truth, we should note that there are limits to our ability to know truth. Drawing from Kant, and Gödel’s incompleteness theorem, there are things we just can never know the truth of.

So I feel that ‘what I currently know to be true’ is not the same as ‘what is true.’ But maybe there might be good reason to argue that ‘what I can possibly know to be true’ should be the same as ‘what is true.’ This requires a rejection of the inaccessible absolute Truth, in favor of an accessible but subjective truth. Reminds me of Ivan Karamazov’s “Even If” in Dostoevsky’s TBK.